i948
MAXWELL GORDON, JOHN 6. MILLER AND ALLANR. DAY
VOl.
70
type; furthermore, it is difficultto find a common factor between nitrates and malonates apart from their low activity coefficients. I wish to thank Professor N. S. Bayliss for his kind permission to use the facilities of his department.
between salts of the type COOH(CH&COONa and CH3(CH&K!OONa, a difference which is marked in spite of the anomalous behavior of sodium caproate6 a t high concentration (probably due to micelle formation). Along with this difference between the two types of salts are to be noted the very low values of the activity coefficients of the salts now under investigation, suggesting that the hydrogen malonate, succinate and adipate ions belong to the same category as the nitrate, chlorate and primary phosphate ions. Thus, although “localized hydrolysis” may be a factor resulting in decreasing dispersion of the curves of the sodium and potassium salts as we proceed from the malonate to the adipate, there must be another factor of larger magnitude which lowers the curves of both the sodium and the potassium salts. It is difficult to explain this effect for the nitrate ion and it is equally difficult to understand why it should occur with ions of the hydrogen malonate
Isopiestic measurements have been made on solutions of the primary sodium and potassium salts of malonic, succinic and adipic acid and the osmotic and activity coefficients calculated. Although these salts are derived from weak acids, their activity coefficient curves are in the order Na > K; in this respect they differ from the sodium salts of monobasic fatty acids. Moreover, their activity coefficients are very low and have an unexpected resemblance in this respect to sodium and potassium nitrate and the primary phosphates.
(6) E. R. B. Smith and R. A. Robinson, Trans. Foroday SOC.,88, 70 (1942).
NEDLANDS, WESTERNAUSTRALIA RECEIVED FEBRUARY 4,1948
[CONTRIBUTION FROM THE
Summary
DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING OF THE UNIVERSITY OF PENNSYLVANIA]
Effect of Structure on Reactivity.l I. Ammonolysis of Esters with Special Reference to the Electron Release Effects of Alkyl and Aryl Groups BY MAXWELL GORDON,~ JOHN G.MILLER AND ALLANR. DAY In the interest of augmenting present knowledge of the effect of structure on reactivity, it was decided to measure the rates of ammonolysis of several homologous series of esters. Although several reports have appeared in the literature on the effect of structural variations on the rate of ammonolysis of esters, they have embraced only limited numbers of esters, the possibility of ester interchange in working in alcoholic solution was often overlooked, no calculations of rate constants or activation energies were made in many cases, and the conclusions derived from these studies were, in some instances, conflicting. Most of the references in the literature on this subject have been discussed by Gorvin.a Additionally it has been shown by Komatsu and Nakayama4 that ethyl phenylacetate reacts with ammonia fifteen times as rapidly as does ethyl benzoate. The present study of the ammonolysis of esters was undertaken in an effort to show the effect of various R and R’ groups in the ester RCOOR’, with special emphasis on tracing the electron release effects in various homologous series. Most (1) From a thesis submitted by M . Gordon to the Department of Chemistry and Chemical Engineering of the University of Pennsylvania, Dec. 1947, in partial fulfilment of the requirements for the degree of Doctor of Philosophy. (2) National Institute of Health Predoctoral Research Fellow, 1946-1948. (3) Gorvin, J . Chcm. Soc., 732 (1945). (4) Romatsu and Nakayama, J . Chcm. SOC.( . ~ a p n n )64, , 558-569 (1933)
of the mechanisms which have been postulated for ammonolysis or hydrolysis of esters6 indicate that the reactivity of the ester depends on the polarization of the carbonyl group in the ester which then provides the center for nucleophilic attack a t the carbon of the carboxyl group. From this representation it may be seen that the greater
“-
R d
cOtR’
the electron release of the R or R’ group, the slower should be the rate of ammonolysisof the ester since the positive charge on the carbonyl carbon would be reduced in magnitude. One may consider two possible mechanisms for the ammonolysis of esters (I)5s6
F;-
R--C-OR’
1
R
+ :NH1 q
i OR’]
LH+I~H*
+R C O N H s + R‘OH
J
Betts and Hammet‘’ suggested that both the am(5) H. Meyer, Monatsh., 47, 31 (1906); Holmberg, B n . , 46, 2997 (1912); Ingold and Ingold. J . Chcm. Soc., 758 (1932); Polanyi . SOC.,80,508 (1934): Day and Iagold, ibid., and Szabo, T ~ o n sForod. 87, 689 (1941); Watson, ibid., 87, 712 (1941). ( 6 ) Chattaway, J. Chcm. Soc., 355 (1936). (7) Betts and Hammett, THIS JOURNAL, 69, 1568 (1937).
May, 1948
AMMONOLYSIS OF ESTERS WITH ALKYLAND ARYL GROUPS
monia amide ion and molecular ammonia are the attacking reagents in ammonolysis.
r n-
1947
components, which means adjusting concentrations so that the entire series of reaction mixtures is kept homogeneous; (3) stability of ammonia solutions, which involves addition of sufficient water to avoid loss of ammonia in the course of the manipulations of the study. Since the above three factors are more or less mutually exclusive, it was only on the basis of a great deal of trial and error that a satisfactory set of conditions was arrived at that would hold for substantially all of the esters employed.
Both mechanisms involve the addition of a nucleophilic agent to the carbonyl carbon atom. The The dioxane used, a product of Carbide and Carbon fact that the presence of a polar (ionizing) solvent Chemicals Corp., was purified by distillation from sodium Alcohols used were freed from water by the is necessary for ammonolysis to occur readily sug- metal. method of Lund and Bjerrum.lo Most of the esters used11 gests that (11) is more probable than I. It is not were available commercially with the exception of methyl al- isobutyrate, methyl trimethylacetate, n-propyl benzoate, necessary to postulate free amide ions ("2)though they may exist to some extent. It is pos- t-butyl benzoate, t-butyl acetate, P-naphthyl acetate, and lactate. These esters were synthesized by a variety sible that through hydrogen bonding the polariza- t-butyl of methods, all of which have been described in the literation of the H-NH2 bond may be increased to the ture. Of special interest is the method of Richard12 for point that it simulates ionization. esterifying sterically hindered acids. This procedure is
Experimental It was originally hoped to carry out the ammonolysis of esters in anhydrous media, but preliminary experiments showed that ammonolysis of most esters will not proceed a t a useful rate in anhydrous dioxane or methanol. This result c o n k n s the work of Grant and Hinshelwood*and other^.^ As Gorvin* has noted, aqueous ammonia is not suitable for accurate kinetic studies owing to the complicating factors introduced by the low solubility of the ester. This difficulty may be overcome by the use of suitable organic diluents in the reaction mixtures. The principal solvent employed in this study was 1,4-dioxane. This compound was used for all series of esters except where all of the esters of a series contained the same alcohol component, in which case that alcohol was used as the diluent. Alcohol was used as a diluent where possible because ammonolysis proceeds roughly twice as rapidly in primary alcohols as in dioxane, provided water is present in each case. Dioxane was selected for use as a result of a process of elimination. Alcohols, generally speaking, were unsatisfactory due to the possibility of ester interchange. There was no method immediately evident for ascertaining the extent of this side reaction. Other solvents react with either the esters or ammonia, or interfere with the subsequent titrations of the reaction mixtures, or are not soluble in water. In the establishment of conditions for this study three major experimental factors had to be considered: (1) the kinetics of the reaction, which involves making the rate of reaction great enough to minimize the experimental errors involved in the measurements; (2) miscibility of the (8) Grant and Hinshelwood, J . Chcm. SOL,1351 (1933). (9) Very recent work in this laboratory indicates that by working
in ethylene glycol and related compounds, amides can be obtained in good yield under anhydrous conditions. This material is part of the
next psper of this aerirs whiah will be rrsdr for publiestha in the .w f u t w ,
seldom encountered in contemporary chemical literature. An azeotrope of methanol and methyl trimethylacetate was encountered which is believed not t o have been hitherto described. Its composition was 12.8% of ester by weight; b. p. 65.0" at 760 mm. The esters, whether synthesized or obtained commercially, were dried over magnesium sulfate and fractionated in all-glass apparatus using an efficient column. The high boiling esters were distilled under reduced pressure. Refractive indices were used t o ascertain final purity of esters employed. Boiling points or melting points and refractive indices (Dline) of the esters used were as follows: methyl acetate, b. p. 57.1'at 755mm., 1.3610; ethyl acetate, 77" at, 760 mm., 1.3728 at 20'; n-propyl acetate, 101-101.5 a t 750 mm., 1.3844,?0'; isopropyl acetate, 88.5-89.0" a t 750 mm., 1.!771,20 ; n-butylacetate, 126.5'at 750 mm., 1.3938at 20 ; isobutyl acetate, 116.5-117.0' a t 750 mm., 1.3900 at 20'"; s-butyl acetate, 112.0-112.5" at 755 mm., 1.3885 a t 20, ; t-butyl acetate, 97.5-98.0' a t 755 mm., 1.3860 a t 20 . n-amyl acetate, 148.0-148.5' at 755 mm., 1.4025 a t 20 ''; benzyl acetate, 93 ' at 7 mm., 1 ;5020 at 20"; phenyl acetate, 78" at 10 mm., 1.5033 at 20 ; vinyl acetate, 69-70" at 75: mm., 1.3941 at 20"; or-naphthgl acetate, m. p. 44-45 ; @-naphthylacetate, m. p. 68 methyl benzoate, 85" at 10 mm., 1.51zO at 20'; ethyi benzoate, 90" a t 10 mm., 1.5058 at 20 ; nipropy1 benzoate, 101.0-101.5' at 9 mm., 1.5000 at 20 ; isopropyl benzoate, 94"at 10 mm., 1.49407 a t 20'; t-butyl benzoate, 10:' a t 16 mm., 1.4896 at 25 ; phenyl benzoate, m. g. 70 ; benzyl benzoate, 110' at 0.5 mm., 1.5685 at 20 ; methyl formate, 31.5"at 760 mm., 1.3438 a t 20"; methyl propionate, 79' a t 750 mm., 1.3770 a t 20'; ?ethyl isobutyrate, 92.0-92.5' a t 755 mm., 1.3840 at 20 ; methyl trimethylacetate, 99.5-100.5" at 760 mm., 1.3908 at 20'; methyl crotonate, 119-120" a t 755 mm., 1.4250 at 20"; methyl phenylacetate, 79" a t 5 mm., 1 . y 9 0 a t 16'; methyl lactate, 50' at 17 mm., 1$40 at 20 ; ethyl lac; tate, 54" at 12 mm., 1.40131 a t 20 ; n-propyl lactate, 71 a t 20 mm., 1.!170 at 20 ; isopropyl lactate, 51 'at 8 mm., 1.4102 a t 20 ; n-butyl lactate, 77' a t 13 mm., 1.4215 a t 20 ; isobutyl lactate, 58" at 3.5 mm., 1.4184 at 20"; s-butyl lactate, 42' a t 1 mm., 1.!168 a t 20'; t-butyl lactate, 40" a t 6 mm., 1.2085 at 30 ; n-amyl lactate, 66" at 1 mm., 1.4261 at 20 ; allyl lactate, 55" a t 3 mm., 1.4369 at 20,"; n-butyl hydracrylate, 69" a t 0.5 mm., 1.4288at20 .
-
(IO) Lund and Bjerrum, Bcr., 64, 210 (1931). (11) We are indebted to Dr. C. H. Fisher of the Eastern Regional Laboratory, U. S.Department of Agriculture, for generous eamplmi Of lactic acid esters. (18) Riohard, AYR, d i m , p k y r i , si, aPa-406 C l B i O ) ,
MAXWELL GORDON, JOHN G. MILLERAND ALLANR. DAY
1948
1'01. 70
TABLE I
BIMOLECULAR REACTION RATESFOR AXUONOLYSIS OF ACETICACID ESTERSAT 25" Time, hr.
Ammonolysis,
%
AmRate Const.
Acetate -+ Methyl 100 33.0 0.00148 200 50.0 .00142 6 2 . 0 .GO140 300 72.0 .00138 400 500 80.6 .00140
monolysis,
%
Rate const.
Ammonolysis,
5%
Rate const.
Ammonolysis,
%
Am-
Rate const.
Ethyl n-Propyl Isopropyl 13.4 O.OOO530 12.1 0.000504 5 . 2 0.000231 22.0 . O W 5 0 19.2 .000436 7 . 5 .OM202 28.1 .OOO420 24.6 .Om370 11.5 .000180 32.6 .000390 2 8 . 5 .000302 14.5 .000156 36.0 .000380 31.6 .000240 16.0 .mol35
Acetate 3 s-Butyl t-Butyl n-Amyl Benzyl 100 5 0 0 000146 4 . 0 0.000111 7 . 0 0.000258 21.9 0.000960 000122 4 . 0 .000100 11.0 .000230 36.3 .000950 200 5.5 5.9 000101 4 . 5 .000090 13.0 ,000205 48.0 .000950 300 .000080 16.3 .000179 58.0 .000950 6 0 000090 5 . 3 400 .000068 18.1 .000154 67.4 .000950 .000080 5 . 3 500 6 2 Phenyl acetate," 0 . 5 hr., ammonalysis 86.6%, const. 2.02 Vinyl acetate: 0 . 5 hr., ammonalysis 75.3%, const. 1.34 0 Ammonolysis was too rapid for a satisfactory rate study, time, 0.5 hour.
In carrying out the ammonolyses, equivalent molar quantities of the various esters were measured into 10-ml. glass ampoules by means of a 1-ml. glass syringe graduated in 0.01 ml. subdivisions and fitted with a #16 gage 3-inch stainless steel needle. The use of the above device made possible the introduction of the esters into the ampoules without wetting the necks of the containers. It was found by a series of control experiments that the quantities of ester delivered by this method were reproducible with an error of less than one-tenth of one per cent. I n the case of the volatile esters the opening of the ampoule was tightly closed by means of half a gelatin capsule of appropriate size, then the needle was pushed through the top of the capsule into the ampoule. In all cases the ampoules were closed off by means of intact capsules after the introduction of the esters and were then chilled in ice. Reproducible results were obtained in the manipulation of the aqueous dioxane-ammonia solutions only by filling eight ampoules from each 100-ml. buretful of solution, using the first and last 10 mI. of each buretful as controls. These control portions of ammonia solution were run into 4% boric acid solution and titrated directly with halfnormal hydrochloric acid using methyl red as the indicator. Following the introduction of both ester and ammonia the ampoules were again cooled in ice and then rapidly sealed off in a hot flame, first removing the gelatin closure. This chilling procedure cut down the loss of ammonia during the sealing operation and reduced the likelihood of leaks developing in the sealed amDoules. The ampoules were then shaken thoroughly and placed in a constant temperature bath. A series of duplicate reaction mixtures was started for each ester. At intervals the ampoules were removed from the thermostat and again chilled. The ampoules were then opened, rinsed into 4% boric acid solution, and titrated directly with standard acid as in the case of the blank runs. The accuracy of the method was ascertained by setting up a series of reaction mixtures and titrating them at intervals as before, but omitting the ester. These blank runs showed that the error in the rest of the method was less than that involved in titrating the solutions. The amount of competing hydrolysis occurring in the ammonolysis of the acetates was determined by the method of Pucher's as modified by F r e n ~ h . 1 Result? ~ of this direct analytical method corroborated the results obtained in estimating ammonia consumed by difference. (13) Pucher, Vickery and Leavenworth, i n d . Eng. Chcm., Anal. Ed.. 7, 152 (1935). (14) French, Johnson and Ratekin. THIS JOURNAL, 18, 1346 (1936).
monolysis,
%
Rate const.
Ammonolysis,
%
Rate const.
n-Butyl Isobutyl 7 . 2 0.000273 5 . 1 0.000201 11.3 .000240 7 . 0 .OW166 13.6 .000209 9 . 5 .000153 16.0 .000178 11.1 .000136 17.3 .000146 14.0 .000128 a-Naphthyl 1.6 0.000070 2 . 5 .000065 3 . 2 .000060 3.8 .000054 4 . 2 .000049
,%Naphthyl 1 . 9 O.ooOo82 2 . 8 .oooO70 3 . 2 .oooO60 3 . 7 .000050 4 . 1 .000046
Determination of activation energies for methyl acetate and the lactates was carried out by setting up parallel reaction mixtures of the various esters and running half of them a t 25' and the rest at 30 or 35". Determination of the activation energies for the benzoates, for the methyl esters other than acetate and formate, and for the acetates above methyl, was omitted because their low reaction velocity made the accurate measurement of temperature coefEcients impossible. The ammonolysis of methyl formate proceeded too rapidly for the determination of activation energies with any accuracy.
Sources of Error A major consideration in working with aqueous ammonia solutions is the possibility of simultaneous hydrolysis and ammonolysisof the ester. This matter will be discussed after presentation of the data. The possibility of reversibility of the reaction was eliminated when it was found that acetamide will not react with alcohols in aqueous dioxaneammonia solutions at the temperatures of the experiments. Other side reactions such as the formation of amines or double amides have been ruled out by various investigators. This aspect of the problem has been fully discussed by Betts and Hammett.? The possibilities for manipulative error were discussed in the description of the experimental method employed, and it is believed that the maximum error for any single determination is about 2%. The over-all errors persisting in any of the rate constants obtained graphically probably amount to less than 5%.
Results The values found for the second order rate constants in the ammonolysis of esters of acetic acid a t 25' are listed in Table I. Also tabulated are the percentages of total reaction for the times indicated. The amount of ester employed in all of the reaction mixtures of this investigation was 0.007 mole per 10 ml. reaction vessel, or 0.7 mole per liter. The ammonia concentration in every
May, 1943
XMMONOLYSIS OF
ESTERSWITH ALKYLAND
I949
L~RYL GROUPS
TABLEI1 Table I1 lists the relative rate constants for the RELATIVE REACTION RATES FOR AMMONOLYSIS OF ACETIC acetates. All of the rate constants have been divided through by the corresponding rate for methyl ACIDESTERSAT 25"
acetate so that the latter always has the value unity in the tabulation, and the comparative rates of the other esters may be seen accordingly, The ammonolysis of benzoic acid esters proceeded so slowly that, in most cases, a satisfactory kinetic study was impractical. The differences in ammonia content found on successive titrations over a period of several days were often so small as to be within the limits of experimental error of the method. Accordingly, only one titration was carried out on each ester after reacting with aqueous dioxane-ammonia for 618.5 hours. Dioxane was used as the diluent. Bimolecular rate constants and comparative rates of ammonolysis of the benzoates are given in Table 111. Second order rate constants and percentages of total reaction in the ammonolysis of lactic acid esters a t 30' are listed in Table IV. The diluent for this series was dioxane. TABLE 111 The values found for the second order rate conBIMOLECULAR REACTION RATESFOR AMMONOLYSIS OF stants and the percentages of total reaction in the BENZOATES AT 25 ammonolysis of the methyl esters of various acids Comparative a t 25' are listed in Table VI. The diluent used for Eater Rate constant rate this series was absolute methanol. Phenyl benzoate" 0.0620 376 The compilation in Table VI11 is taken from Methyl benzoate .000165 1.00 Tables 11,I11 and V in an effort to determine what Benzyl benzoateb .000087 0.527 correlation may be found between structure and Etkyl benzoate .000084 .509 reactivity. n-Propyl benzoate .000065 .391 The extent of ammonolysis and hydrolysis of Isopropyl benzoate ,000056 .341 acetic acid esters is shown in Table IX. $-Butyl benzoate .000019 .116 From the temperature coefficients of ammonolyOwing to the high molecular weight of this ester only 0.467 mole/l. was used. Any greater amount led to separa- sis an approximate evaluation of the activation tion of layers. Reaction time in this case was twenty-four energies was possible using the integrated form of hours. b 0.467mole/l. of this ester was used. the Arrhenius equation. Calculations of PZ faccase was between 2.5 and 2.8 moles/liter. All of tors from the equation k = PZe-AIRT showed the reaction mixtures contained ten moles of wa- that errors in the values of activation energies were ter/liter. The diluent used for this series was 1,4- too great to permit the use of entropy changes to dioxane. The rate constants are bimolecular and show the validity of structure-reactivity correlations. are expressed in liters/mole/hour. 100 hr.
Ester
200 hr.
300 hr.
1422 1443 1365 Phenyl acetate" Vinyl acetate" 909 944 957 Methyl acetate 1.00 1 .00 1.00 Benzyl acetate 0.649 0.662 0.678 Ethyl acetate .358 ,317 ,300 .341 .307 %-Propylacetate .264 .169 .149 %-Butylacetate .185 ,174 .163 .148 n-Amyl acetate .156 .142 .128 Isopropyl acetate ,136 .117 .lo9 Isobutyl acetate s-Butyl acetate .0986 .0859 .0721 t-Butyl acetate .0750 .0655 .0643 @-Naphthylacetate .0554 .0493 .0429 a-Naphthyl acetate .0473 .0458 .0429 Reaction has gone to completion by the time 100 hours is reached, but this ester is included for purposes of comparison.
O
TABLE IV BIMOLECULAR REACTION RATES FOR AMMONOLYSIS OF LACTIC ACIDESTERS AT 30" Time, hr.
Lactate + 10 20 30 50 100 150
Ammonolysis,
%
Rate const.
Methyl 55.0 0.0397 78.0 .0372 88.0 .0363 94.0 .0361
Ammonolysis,
%
Am-
monoRate const.
lysis,
%
Rate const.
Allyl Ethyl 54.0 0.0370 25.9 0.0155 77.5 .0370 37.0 .0135 88.0 .0362 45.2 .00950 57.8 .00800 78.5 .00780
Lactate ---f n-Amyl Isobutyl Isopropyl 30 32.5 0.00504 28.2 0.00493 11.1 0.00242 50 42.0 .00500 38.3 .00483 16.0 .00188 100 60.9 .00497 57.5 .00460 25.2 .00137 150 76.0 .00482 71.0 .00440 31.3 .00108 Inserted for comparison with n-butyl h t a t e .
Ammonolysis
%
Rate const.
Ammonolysis,
%
Rate const.
Ammonolysis,
%
Rate const.
+Butyl n-Propyl 16.6 0.00780 26.1 .00760 33.6 .00740 34.0 .00740 45.0 ,00710 46.0 ,00710 66.2 .00610 66.1 .00608 75.1 . a 7 0 n-Butyl :-Butyl hydracrylate' s-Butyl 11.0 0.00200 1.10 0.00042 8.5 0.00167 16.1 .00184 1.79 .O0040 12.0 .00140 21.0 .00132 18.4 .00100 24.9 .00087 24.0 .o0080
MAXWELL GORDON, JOHN G. MILLERAND ALLANR. DAY
1950
VOl. 70
TABLE V methyl acetate and the lactates are listed in RELATIVE REACTIONRATES FOR AXMONOLYSIS OF LACTIC ACID ESTERSAT 30O
x'
30 hr.
Ester
Discussion
50 hr.
While the comparative rates for similar esters of different acids (Table VIII) show that the correlation of structure and reactivity is still far from quantitative, they do show that the relative magnitude of effects may be determined. The variations in reaction rates of most of the esters studied are a t least qualitatively predictable from the known electronegativity of the groups attached. While it is true that probably both steric and polarization effects contribute to the variations in rate constants among the reactions of this study, the fact that, except for the case of the naphthyl
Methyl lactate 1.00 1.00 Allyl lactate 0.997 Ethyl lactate .262 0.221 n-Propyl lactate .204 .197 n-Butyl lactate .204 .197 n-Amyl lactate .139 .139 .136 ,134 Isobutyl lactate .0667 .0.521 Isopropyl lactate +Butyl lactate .0551 .0509 n-Butyl hydracrylate" .0460 .0398 .0116 .0111 &Butyl lactate " Tabulated for comparison with n-butyl lactate.
TABLE VI BIMOLECULAR REACTION RATESFOR AMMONOLYSIS OF METHYLESTERSAT 25" Time, hr.
Ammonolysis, %
Methyl +
50 100 150 200 300
Rate const.
Ammonolysis, %
Formate
Rate const.
Ammon-
olysis, %
Lactate
79.3" 1.54
94.0b
Methyl --+
Acetate 45.0 0.00500 63.5 ,00480 74.9 .00453 81.3 .00420
0.0361
Propionate
25.0 36.8 46.0
Rate const.
Ammonolysis, %
Rate const.
Ammonolysis, %
Phenylacetate
40.0 59.3 71.0 77.0 84.2
Benzoate
0.00469 .00435 .00408 .00375 .00315
Crotonate
29.0 52.8 71.3 82.6
Isobutyrate
16.0 18.3 20.8 23.1 27.7
methanol the rate is approximately twice this value. Time, 388 hours.
0.00290 .00332 ,00370 .00403
Trimethylacetate'
0,00188 10.0 0.000880 50 13.3 .OW725 .00110 100 16.6 .00075 .000600 150 54.0 .00058 20.0 .000500 200 .00043 25.0 .000390 67.4 300 This rate is for Ammonolysis is too rapid for a satisfactory rate study; time, 0.5 hour. 0.00282. .00233 .00196 .00173 .00160
Rate const.
0 0 0 0 1.43d
0 0 0 0 .0000147
dioxane diluent. In Ammonolysis is too slow for a satisfactory rate study.
TABLE VI1 TABLE VI11 RATESOF SIMILARESTERSOR DIFFERENT RELATIVE REACTION RATESFOR AMMONOLYSIS OF METHYL COMPARATIVE ESTERSAT 25" ACIDS Ester
50 hr.
100 hr.
200 hr.
321" 367" 308" Methyl formate 7.56 7.9b 7.2 Methyl lactate 1.00 1.00 1.00 Methyl acetate Methyl phenyl0.906 0.893 0.938 acetate .692 .881 Methyl crotonate' .so .412 .485 .564 Methyl propionate .138 .239 .376 Methvl benzoate Methyl isobutyrate .176 .151 ,119 Methyl-trimethylacetate 00294d .00306.j .00350.j This ratio is obtained from the rate constant after onehalf hour and is used here for purposes of comparison. b Taken from data for 50 hours. The reaction of this ester results primarily in l14-addition of ammonia. d Taken from data for 388 hours.
.
Possible errors in activation energy calculations were estimated by substituting probable maximum and minimum values of the various rate constants in the integral forin of the Arrhenius equation. A&%h&&dQn W W ~ h bet the ammQnQ1ySiEi Qf
Alcohol component
Acetic acids
Benzoic acid b
Lactic acid C
Methyl 1.00 1.00 1.00 Ethyl 0.358 0.509 0.221 %-Propyl .341 .391 .197 Isopropyl .156 .341 .0521 &Butyl .0750 .116 ,0111 "This comparison is for a reaction time of 100 hours. Reaction time is 618.5hours. Reaction time is 50 hours.
esters to be discussed later, an explanation based on polarization effects has been found to be adequate has led to the emphasis on the latter. The relative rates for the ammonolysis of the different esters agree in the main with published values obtained in alkaline and acid hydrolysis of esters.I6 In all of the ester series where the acid component was held constant, we noted that the reactivity of the ester toward ammonia decreased with increase of the molecular weight of the alcohol (15) Skrabal, Monatsh., 46, 148 (1924). el seq.; Palomaa, B e t . , TlB, 480 (1938); Khdlcr, Bet., 8 0 8 , P7P2 (10.96); Sudbn*su8h, beClrr*l, TBr 407 {IeQQ).
?Jay, 1948
=\MMONOLYSIS OF
ESTERSWITH ALKYLAND ARYLGROUPS
1951
0TABLE IX EXTENTOF AMMONOLYSIS AND HYDROLYSIS OF ACETATES V ~ = C ! ! - - C H I % % Am0 0 Ester Hydrolysis monolysis Time, hr. II Methyl acetate 1.1 75.9 579.5 a - O - C - C HII s f--f ( - >=&-C-CH, L 1.1 78.6 699.5 Methyl acetate (3 forms) 0.7 83.9 792.5 Methyl acetate on the reactivity of phenyl acetate is the same as 1.1 35.4 698.5 Ethyl acetate that discussed above for vinyl acetate. 4.3 26.0 862.5 n-Propyl acetate On the basis of resonance it might further be an3.3 7.6 1127 Isopropyl acetate ticipated that the naphthyl acetates would have 3.6 23.9 1127 n-Butyl acetate similarly enhanced reactivities. However, these 4.4 17.1 1127 Isobutyl acetate esters react very slowly with ammonia. This 2.9 3.6 1127 s-Butyl acetate phenomenon can be partly attributed to the large 2.9 2.9 1127 &Butyl acetate size of the naphthyl groups. Atomic models in4.3 20.3 1127 a-Amyl acetate TABLE X ACTIVATION ENERGIES FOR AMMONOLYSIS Activation energy, cal./mole
Ester
Methyl acetate Methyl lactate Ethyl lactate n-Propyl lactate Isopropyl lactate *-Butyl lactate Isobutyl lactate s-Butyl lactate n-Amyl lactate Allyl lactate
12,700 11,800 10,500 10,700 9,900 9,400 11,800 11,800 10,200 11,100
* 2000 * 1000 * 1000 * 1000
* 1000 * 1000 * 1000
f
* z= !
1000 1000 1000
component, and, to an even greater extent, with increased branching of the alcohol. Generally speaking, esters of primary alcohols ammonolyze more readily than those of secondary alcohols, and the latter in turn react with ammonia more rapidly than the esters of tertiary alcohols: CHa > CzHs > CHsCHaCHl > (CH&CH > (CH&C. These results are in accord with accepted electron release effects of alkyl groups. In the acetate series we found that vinyl acetate had several hundred times the reactivity of ethyl acetate. Writing the resonance forms for vinyl acetate, we see that the positive charge on the oxygen tends to make the carbonyl carbon more CHt=CH&
.t+ -CHrCH=O+-
electrophilic, which in turn greatly enhances the reactivity of this carbon toward a nucleophilic 0 -CH*-CH=O+
I
c &CHa
attacking species. The latter, in this case, is probably either the ammonia amide ion or a hydrogen bonded waterammonia complex. The very great reactivity of phenyl acetate toward ammonia, compared to the other acetates, is likewise attributable to resonance, in this case mainly between the benzene ring and the oxygen bound to it. Writing the various resonance forms we see that all of the charged structures have a positively charged center adjoining the carbonyl carbon. The &ect of thia charge concentration
dicate the reduced susceptibility of the carbonyl carbon to attack due to steric hindrance. In the forms of phenyl acetate in which there is resonance with the benzene ring, there is the spatial requirement that the carbonyl carbon be coplanar with the ring. The decreased likelihood of this restriction being complied with in the naphthyl acetates may further account for their low reactivity toward ammonia. It was noted that benzyl acetate arnmonolyzed twice as rapidly as ethyl acetate and allyl lactate reacted with ammonia five times as rapidly as did n-propyl lactate. From these results it must be concluded that the benzyl group is less electron releasing than the ethyl, and the allyl less than the n-propyl. These relationships would seem to be a t odds with the well-known behavior of the halides of these compounds. For example, benzyl chloride is known to be more reactive than ethyl chloride and this phenomenon is popularly attributed to the greater polarization of the carbonhalogen bond in the benzyl chloride, from which one would have to conclude that the benzyl group is more electron releasing than the ethyl. Similarly the greater reactivity of allyl chloride over npropyl chloride would lead one to believe that the allyl group is more electron releasing than the nPropyl. However, reference to dipole moments shows that precisely the reverse is true. The relative dipole moments of ethyl bromide16 and benzyl bromide” are 2.12 (in benzene) and 1.85 (in benzene), respectively, indicating that the carbonhalogen bond is less polarized in the latter, and that the electron release effect of the benzyl group must be less than that of the ethyl group. Similarly the relative dipole moments of n-propyl bromide16 and allyl bromidels are 2.00 (in benzene) and 1.79 (in benzene), respectively, again leading to a conclusion supporting the results of this investigation, namely, that the allyl group is relatively less electron releasing than the n-propyl. The usual high reactivity attributed to allyl halides, alcohols, etc., in unimolecular displacements, when compared to the corresponding saturated (16) Daily, Phys. hu., [iil 84, 648 (1929). (17) Smyth and Wsflr, T a r n JOURNAL, 64,1884 (111) Rsrtrl Pa #bY** C k r w ~B11, 812 (1981).
(i~ai).
1952
MAXWELL GORDON, JOHN G. MILLERAND XLLANR. DAY
Vol. 70
alkyl compounds, is due to stabilization of the obtained by varying the R’ groups as discussed carbonium ion through resonance. However, earlier. However, methyl trimethylacetate has a only after ionization does this stabilizing influence much lower rate, compared to methyl acetate, than t-butyl acetate has compared to methyl CH2=CH--CH2-X ---f CHFCH-CH~’ Xacetate. The greater retardation effect of the R3C grouping in methyl trimethylacetate can be at‘CHz-CH=CH* tributed to the closer proximity of that group to operate, so that while the allyl compound is in the the carbonyl carbon than is the case in t-butyl aceun-ionized form it is perfectly consistent to say tate. The oxygen, due to its unshared electrons, that the polarization of the carbon-halogen bond in is probably an efficient conductor of inductive forces, but since the inductive effect falls off rapallyl halides is lower than in n-propyl halides. There is abundant evidence in the literature, idly with distance, replacing a given R group in the well summarized by Branch and Calvin,’Qto show ester RCOOR’ by any given radical will have the great effect of resonance stabilization by cat- greater effect on the rate of ammonolysis of the ions such as allyl and benzyl groups in bimolecular ester than replacing the corresponding R’ group by the same radical. (SNz, displacements In methyl benzoate, unlike phenyl acetate, the The fact that the reactivity of benzyl halides is greater than that of ethyl halides is due to the conjugation of the carbonyl group with the benpresence of the allylic system in the former. So zene ring decreases the reactivity toward amwe see that the inhence on reactivity attributed monia. It should be noted that the oxygen atto the allyl and benzyl groups in this investiga- tached to the methyl group of methyl benzoate is tion is not incompatible with the normal behavior less positive than that of methyl acetate due to of these compounds in displacement reactions, resonance with the ring in the former ester. Writsince in ammonolysis there is no cleavage of the ing the resonance structures of methyl benzoate we can see the reasons for the reduced electroalkyl-oxygen bond. In the case of the saturated aliphatic esters of 0benzoic and lactic acids the same generalizations apply as in the case of the acetates. The high a reactivity of allyl lactate, as compared to n-pro0 0pyl lactate, was noted above. The rate for vinyl O - C = O -II- C H I E = = C - OI - C H ~ esters is much greater than that of allyl esters due to the absence of resonance in the un-ionized form b c (3 forms) of the latter. n-Butyl hydracylate is less than philic nature of the carbonyl carbon atom. one-fgurth as reactive as n-butyl lactate because the electron attracting hydroxyl group is farther Methyl acetate has only the forms corresponding removed, in n-butyl hydracrylate, from the reac- to a and 6 above and would therefore be expected tive center of the molecule. In general, all of the to ammonolyze more rapidly than methyl benzolactates have from ten to twenty times the reac- ate. In the reaction of methyl crotonate Morsch*O tivity toward ammonia that is exhibited by the corresponding acetates. Here again the difference reported predominantly l,4-addition of ammonia is due, in part, to the presence of an electron at- to give methyl 3-aminobutyrate, which he isotracting group in the lactic acid esters. Further- lated. In this laboratory these results were conmore, to the extent that it occurs, hydrogen firmed by titrating starting mixtures and reaction bonding in the lactates and hydracrylates could products by the bromate-bromide method, contribute to the electrophilic activity of the car- whereby about 96% of the product was found to bonyl carbon atom since it assists in the polariza- be saturated. These results probably account for tion of the carbonyl carbon-xygen bond. How- the rise in reaction rate constants observed with ever, hydrogen bonding of the hydroxyl hydrogen this ester, as compared with the downward drift to the carbony1 oxygen probably reduces the in- in the purely ammonolytic reactions. In the ammonolysis of methyl formate it was ductive effect of the hydroxyl group, so that the over-all contribution of hydrogen bonding to the found that the competing hydrolytic reaction may account for as much as 40% of the ester con+ CHs-CH c C-0-R sumed, but nevertheless the ammonia reaction +. was still several hundred times as rapid as in the 0-H - case of methyl acetate. Work which had been rate of ammonolysis of the lactates is open to planned with other formic acid esters was dropped due to the great amount of competing hydrolysis. question. The methyl esters of the various acids show that Another objection to the use of formates was the the variation of R groups in the ester RCOOR’ fact that reaction with ammonia was so rapid as to results in a change in reactivity analogous to that make accurate kinetic studies exceedingly difficult. (19) Branch and Calvin. “Theory of Organic Chemistry,” 1941.
5
+
.
82
p. 438.
(a0) Morseh, Monolsh.. SO, 60 (1932):
May, 1948
-1MMMONOLYSIS OF
ESTERSW l r H
The amount of hydrolysis found in the case of esters of acetic acid has been given in Table IX. It should be appreciated that hydrolysis figures presented represent extreme upper limits in view of the length of reaction time. These extreme periods of time were specifically employed because the amounts of hydrolysis obtained over periods of time comparable to those used in the ammonolytic studies were of the same order of magnitude as the experimental error of the method and hence meaningless. Since the extent of hydrolysis in all cases tabulated could not be evaluated for periods comparable to those used in the ammonolytic determinations, we were unable to apply any correction to the calculations for the rates of ammonolysis in order to account for the side reaction. In view of the original objectives of this study, an additional expenditure of time on the further study of the hydrolysis of esters was considered unwarranted. The relative extent of hydrolysis in these experiments, even for extreme periods of reaction time, is well below that encountered by other invest i g a t o r ~owing ~ ~ , ~to~ the lower concentrations of water (18%) employed in this Laboratory. From the foregoing remarks it should be clear that, for the periods of time used in ammonolysis in this study, the amount of hydrolysis is within the limits of experimental error and hence not significant in its effect on the structure-reactivity correlations. The second order rate constants for the ammonolysis of esters have been found to drift downward with time. The only exception was the case of methyl crotonate, in which, as discussed earlier, the principal reaction was not ammonolysis. The downward drift of rate constants is attributable to two causes; first, to the accumulation of ammonium ions from the competing hydrolytic reaction, the ammonium ions acting to retard the (21) Results of unpublished work show that the same qualitative structure-reactivity correlations are obtained by working in anhydrous ethylene glycol-ammonia, in which hydrolysis is improbable, as are reported in this paper for aqeuous solutions. (22) French and Wrightsman, THIS JOURNAL, 60, 80 (1938)
ALKYL .4ND ARYL GROUPS
1953
principal reaction and, second, to the fact that reactions were carried out in concentrated solutions in which we would expect to find considerable deviation from ideality.
Summary The rates of ammonolysis of several series of esters have been determined and have been found to be regulated by the electron release effects of both the R and R’ groups in the ester RCOOR’, with variations of R having the greater effect due to the rapid falling off of the inductive effect with distance. kinetic study has revealed the following order of reactivity of acetates with aqueous ammonia: phenyl > vinyl > methyl > benzyl > ethyl > n-propyl > n-butyl > n-amyl > isopropyl > isobutyl > s-butyl > t-butyl > p-naphthyl > anaphthyl. A study of the reactivity of benzoates indicated the following order: phenyl > methyl > benzyl > ethyl > n-propyl > isopropyl > t-butyl. The following order of reactivity in the ammonolysis of lactates was observed: methyl > allyl > ethyl > n-propyl > n-butyl > n-amyl > isobutyl > isopropyl > s-butyl > t-butyl. By utilizing a series of methyl esters of different acids the effects of structural variations on the acid side of the ester molecule were observed. The following order of ammonolysis of methyl esters was obtained: formate > lactate > acetate > phenylacetate > propionate > benzoate > isobutyrate > trimethylacetate. In general the relative rates of ammonolysis obtained by varying the alcohol and acid components of the ester follow the anticipated electron release effects and confirm the results obtained by other investigators in the acid and alkaline hydrolysis of esters. An azeotrope of methanol and methyl trimethylacetate has been described. PHILADELPHIA, PENNSYLVANIA RECEIVED NOVEMBER 14, 1947
